The OSCE cycle: roles, the 12-month timeline and the master checklist

An OSCE takes about twelve months to plan well and one day to run. The cycle has seven phases: governance, blueprint, station writing, people, logistics, the exam day, and the analysis and results that follow. A named OSCE lead owns the whole cycle. This page gives the roles, a month-by-month timeline, and a master checklist.

Why the cycle matters more than the day

Most OSCE failures are decided months before exam day. A station that tests the wrong skill was approved in a hurry. An examiner who marks two grades harder than the panel was never calibrated. A pass mark that could not be defended was chosen the week before the results. On the day itself, these problems only become visible.

The AMEE Guide on OSCE organization assigns a single person to this role for this reason. It recommends an assessment team with a small sub-committee and a nominated OSCE lead who carries overall accountability, backed by dedicated administrative support. Schools that operate on more than one site add a local team for each site. That structure is the first decision in the cycle, and every later one depends on it.

This page covers the whole year. For the mechanics of the circuit itself, the rotation, the timing signals, and what goes wrong in the room, read the running the exam section. The sections of this guide that follow cover each phase in depth.

Who does what in an OSCE?

An OSCE has more roles than a written exam, and most of them are part-time. The table lists the roles the cycle needs and the decision each one owns. Small schools combine roles in one person. Large schools split them further. The point is that each decision has an owner before the work starts.

Two roles are often missing. The first is the peer reviewer. A station read only by its author reaches the exam with its author's blind spots. The second is the analyst. Without one, the results are ratified without anyone checking whether the exam measured what it was meant to measure.

What happens in each month of the OSCE cycle?

The timeline below is a working schedule for a school that runs one summative OSCE a year. It is built on the seven-step framework in the South African planning study, the eleven-step design sequence in the deployment guide, and the organizational guidance in the AMEE guide. Compress it for a small formative exam. Do not compress the first three months for a high-stakes one.

Three points in this timeline carry more weight than the rest. The blueprint in month two decides what a pass means, so nothing else should start before it is signed off. The pilot in month six is the only chance to find a broken station before candidates do. And the people phase needs more time than schools expect. One estimate in the AMEE guide puts simulated patient training at up to 15 hours per person.

Which decisions must be made before a station is written?

Station writing is the phase schools rush into first, because it feels like the real work. Six decisions need to be made before it starts. Each one is a page in the sections that follow.

1. The blueprint. Which competencies and which problems the exam samples, and how many stations each cell gets. This is the single most important document in the cycle.
2. Station count and length. The AMEE guide notes that stations generally last five to ten minutes. It adds that 14 to 18 stations of that length could reach an adequate reliability for a high-stakes exam. The South African framework used 10 to 20 stations, each lasting 5 to 10 minutes. Decide the number from the blueprint and the evidence, not from the number of rooms.
3. The scoring scheme. Whether a station uses a checklist, a global rating, or both, and how it treats a critical error.
4. The standard-setting method. Chosen before the exam, not after the scores are in.
5. The station template. One template for every writer, with candidate instructions, examiner information, and the simulated patient brief in fixed sections.
6. The review process. Who reads each station, what they check, and who signs it off.

A school that makes these six decisions in months one and two spends months three to five writing stations that fit. A school that skips them spends the same months writing stations that do not, and then discovers it in the pilot or, worse, on the day.

What does the exam office need in the last six weeks?

The last six weeks belong to logistics. The AMEE guide's list of administrative tasks is a good audit. Candidates are allocated to sites and circuits and told how to get there. Paperwork is printed or loaded onto devices. Simulated patients and examiners are confirmed, with reserves for both. The venue has briefing rooms, an office, waiting areas, a quarantine room and refreshments.

Rest stations are placed inside the circuit, never first or last, so that every candidate rests at a comparable point. Spares of every piece of equipment are on site. Three separate briefings are scheduled: one for candidates, one for examiners, one for simulated patients. The deployment guide suggests about 30 minutes for examiner orientation on the day, in addition to the earlier training.

The contingency plan is written now, calmly, not on the day. It answers the questions the floor manager would otherwise improvise answers to: what happens when a simulated patient falls ill, when a device loses its connection, when a candidate arrives late, and when a station's instructions go missing.

What happens after the exam?

The exam is not over when the last bell rings. The AMEE guide lists the after-exam work in order. Mark sheets are cross-checked, and the score spreadsheet is verified. The standard is set, and the results are ratified by the exam board. Results are published. Complaints and appeals are handled. External examiners report. Post hoc psychometrics are conducted, and examiners and candidates are asked to evaluate the exam.

The analysis phase is where the exam earns its next year. Reliability tells the committee whether the exam had enough stations. Station metrics tell the writers which stations to keep and which to fix. Examiner data indicates who drifted. A school that files this data without reading it repeats its faults. A school that reads it improves each cycle.

Where does the cycle usually break?

The cycle breaks in predictable places. Knowing them is most of the defense.

• No owner. Decisions made in corridors cannot be defended in an appeal. Name the lead in month one.
• Blueprint after stations. Existing stations are mapped onto a blueprint, rather than the other way round. The exam then samples what the faculty had, not what the program requires.
• Too few stations. A circuit sized to the available rooms rather than to the reliability the exam requires. The evidence on station count is on the validity and reliability page.
• Untrained examiners. Training reduces examiner variation, as the AMEE guide states, but a morning briefing is not training.
• No pilot. A station's flaws are found by the first candidates instead of by volunteers.
• Standard set after the scores. A pass mark chosen with the results in view is hard to defend and harder to explain.
• No review. The data from this year never reaches next year's planning.

One limit is worth stating. The timeline above assumes a school with a committee, an exam office, and at least one person who can run the analysis. A school starting from nothing needs a simpler first cycle: a formative OSCE with fewer stations, a small volunteer examiner panel, and a review that feeds the first summative exam a year later. The Tunisian implementation study reached the same conclusion: start formative, build capability, then raise the stakes.

The master checklist

Use this list as the agenda for the committee's meetings across the year. Each item links to the section of this guide where the detail is found.

Takeaways

Frequently asked questions

Sources